In the first quarter of 2026, Bitcoin network hash rate experienced its first year-over-year decline since 2020. According to Glassnode data, total network hash rate has dropped about 4% since the start of the year, hovering around 1 ZH/s. This metric had previously maintained double-digit growth for five consecutive years, nearly doubling in 2022. From its peak of approximately 1,160 EH/s in October 2025, hash rate retreated by about 10% to 1,045 EH/s at the end of December. Although it recovered to 1.02 ZH/s by March 2026, the growth momentum has clearly reversed.
Alongside changes in hash rate, hashprice has continued to hit new lows. As of March 28, 2026, hashprice fell 6.65% over three days to 31.60 USD/PH/s/day, after briefly touching 28 USD/PH/s/day earlier in March—the lowest point since the Bitcoin halving.
These shifts are not just short-term market fluctuations; they signal a deeper structural turning point. Over the past five years, Bitcoin hash rate grew nearly tenfold from about 100 EH/s, with increasing miner participation and infrastructure investment. Now, this trend has been broken, indicating that the economic model driving hash rate expansion is changing.
What is the core mechanism behind the hash rate decline?
The fundamental driver of hash rate decline is the systemic deterioration of miner profitability. In the fourth quarter of 2025, the weighted average cash cost for public mining companies to produce one Bitcoin climbed to roughly $79,995, while the price of Bitcoin had already pulled back 31% from its all-time high of $124,500 to $86,000. Entering the first quarter of 2026, Bitcoin price hovered between $67,000 and $70,000, pushing mining operations deep into negative margins—estimated losses per BTC mined reached about $19,000.
A more direct pressure comes from hashprice itself. In Q4 2025, hashprice dropped to the 36-38 USD/PH/s/day range, near the breakeven point for most miners. In Q1 2026, it fell further to 29 USD/PH/s/day. At this level, miners operating mid-generation machines (with energy efficiency around 29.5 J/TH) and paying more than 0.05 USD/kWh for electricity are running at a cash loss. Estimates suggest that 15% to 20% of legacy equipment across the network is operating unprofitably.
Meanwhile, the rise of AI infrastructure offers mining companies an unprecedented alternative. CoinShares data shows that the unit cost for Bitcoin mining infrastructure is about $700,000 to $1 million per megawatt, while AI infrastructure costs $8 million to $15 million per megawatt—a dramatic difference, but AI contracts promise profit margins over 85% and multi-year income visibility. The revenue gap is even clearer at the data center level: AI/HPC data centers generate $200-$500 per megawatt, compared to just $57-$129 for Bitcoin mining.
What are the costs and trade-offs of shifting to AI?
The scale of this transformation far exceeds any previous diversification efforts in the industry. By Q1 2026, public mining companies had announced cumulative AI/HPC contract values exceeding $70 billion. The CoreWeave and Core Scientific agreement is worth $10.2 billion over 12 years; TeraWulf signed a $12.8 billion HPC contract; Hut 8 and Fluidstack entered a $7 billion, 15-year AI infrastructure lease.
However, this transformation is not cost-free. Capital markets are signaling a clear divergence: mining companies with HPC contracts trade at EV/NTM sales multiples of 12.3x, while pure mining firms are at just 5.9x. This premium comes with fundamental changes to balance sheets—IREN took on $3.7 billion in convertible notes, WULF’s total debt reached $5.7 billion, and CIFR issued $1.7 billion in senior secured notes.
Another visible cost is the ongoing reduction in Bitcoin holdings. Public mining companies have sold off more than 15,000 BTC from peak levels to fund AI infrastructure construction. Companies like WULF, CORZ, CIFR, and HUT are essentially evolving into data center operators who also mine Bitcoin. By the end of 2026, AI revenue could account for as much as 70% of public miners’ income, up from about 30% currently.
What does this mean for Bitcoin network security and the hash rate landscape?
A drop in absolute hash rate naturally raises concerns about network security. In theory, a significant reduction in total computational power could lower the cost of launching a 51% attack, potentially affecting investor confidence in the medium term. However, there’s a subtler paradox at play.
US-listed mining companies previously controlled about 41% of global hash rate. As these firms reallocate capital and power resources to AI/HPC operations, their share of Bitcoin mining is shrinking. Nansen Senior Research Analyst Jake Kennis points out that decentralization isn’t determined solely by changes in absolute hash rate—it’s crucial to know who replaces lost hash rate capacity and how ownership is distributed.
A "polarized market" may emerge: large AI-miner hybrid models dominate the high end, ultra-low-cost small operators grow at the margins, and mid-sized pure miners face increasing pressure. Geographically, the US still leads with about 37.5% of global hash rate, but hash rate is spreading to emerging regions like Paraguay, Ethiopia, and Oman, which helps mitigate concentration risk.
Additionally, Bitcoin’s built-in adjustment mechanism is working. Three consecutive mining difficulty reductions (the first since July 2022) signal that "miner capitulation" is underway. Lower difficulty reduces the threshold for remaining miners, marginally improving profitability for those still operating.
How might the mining industry evolve going forward?
CoinShares’ segmented forecasting model suggests that if Bitcoin price rebounds above $100,000, total network hash rate could recover to 1.8 ZH/s by the end of 2026. The validity of this prediction hinges on a key variable—when and whether Bitcoin price triggers a profitability reversal.
A more certain trend is the solidification of industry stratification. Transformation-focused mining companies are rapidly becoming "data center operators who happen to still mine Bitcoin," with AI revenue share steadily rising. Pure mining firms, represented by low-leverage companies like CleanSpark and HIVE, survive by extreme energy cost control and financial discipline. The gap in capital returns between these two models will continue to widen, and capital markets may further differentiate their valuations.
Structurally, Bitcoin mining may shift toward more intermittent and cheaper power sources. AI infrastructure’s rigid demand for stable electricity is raising the competition threshold for premium power resources, while Bitcoin mining’s interruptible load could become a differentiator—filling power gaps during surplus or low-cost periods.
What are the potential risks and limitations?
This transformation path faces several critical risks. First is the high concentration of debt risk. Some miners’ interest expenses have soared from hundreds of thousands per quarter to tens of millions, fundamentally changing the industry’s risk profile. If AI revenue ramps up slower than expected or hash rate demand cycles downward, debt pressure could quickly impact entire balance sheets.
Second is the risk of prolonged low Bitcoin prices. If Bitcoin fails to rebound above $100,000 in 2026, high-cost miners will capitulate faster. Remaining hash rate may be dominated by two types of miners: professional operators with ultra-low electricity costs, and hybrid firms subsidizing mining losses with AI business—where the former ensure network operation, and the latter treat mining as a sideline.
Moreover, the recovery of hash rate from about 920 EH/s to the current 1.02 ZH/s demonstrates both resilience and fragility in the hash rate market. Short-term rebounds are mainly driven by restarting equipment among remaining miners, while new capital investment in mining infrastructure remains suppressed by persistently low hashprice.
Summary
Bitcoin hash rate’s first Q1 year-over-year decline in six years is not an isolated technical fluctuation, but a concentrated reflection of mining industry economic model restructuring. As the gap between production costs and market prices widens and AI demand offers miners a more lucrative outlet for capital, natural capital flows have led to a temporary halt in hash rate growth. The deeper impact is that miners are evolving from "hash rate providers" to "electricity asset management operators," rewriting the industry’s capital structure, valuation logic, and risk profile. For the Bitcoin network, a decline in absolute hash rate does not equate linearly to increased decentralization. The key question is whether, after hash rate loses its previous growth inertia, the cost and ownership distribution of new entrants can foster a more resilient network.
Frequently Asked Questions
Q: Does the decline in Bitcoin hash rate mean the network is more vulnerable to a 51% attack?
A drop in absolute hash rate theoretically lowers the threshold for total computational resources needed for an attack. However, current hash rate remains at a historically high level of about 1 ZH/s, so attack costs are still extremely high. More importantly, hash rate is now spreading from highly concentrated US-listed miners to a broader set of operators, which increases the complexity of coordinating an attack. Network security depends not only on total hash rate, but also on its distribution and the economic incentives of attackers.
Q: Why does hashprice keep falling? What are the main factors?
Hashprice measures miners’ USD income per unit of hash rate and is influenced by three core variables: Bitcoin price, total network hash rate, and the proportion of transaction fees. Since Q4 2025, Bitcoin price has dropped sharply from its $124,500 peak, combined with near-record network hash rate, both squeezing hashprice. Meanwhile, on-chain transaction fees account for only about 0.43% of block rewards, failing to provide meaningful income supplementation. Fundamentally, hashprice reflects the equilibrium price between Bitcoin network hash rate supply and miners’ willingness to pay.
Q: Will miners continue to mine Bitcoin after shifting to AI?
Most transformation-focused miners have not fully exited Bitcoin mining, but have downgraded it to a "parallel business." The core logic is that mining infrastructure—power, land, cooling systems—is highly versatile. After fulfilling AI/HPC contract requirements, remaining power and rack capacity can still be used for Bitcoin mining. This is essentially an optimized allocation of capacity—deploying high-return, stable resources for AI contracts and using surplus capacity to capture potential upside from Bitcoin price rebounds.
Q: Is the hash rate decline a long-term trend or a short-term phenomenon?
According to CoinShares’ forecasting model, if Bitcoin price rebounds above $100,000, hash rate could recover to 1.8 ZH/s by the end of 2026. The key point is that even if absolute hash rate recovers, structural transformation among miners has already occurred—the rising share of AI revenue means hash rate’s sensitivity to Bitcoin price may be permanently altered. Hash rate growth is shifting from "continuous capital inflow" to "price cycle-driven."
